Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $(A_n)$ be a discrete Markov chain with transition matrix $M$ and $B_n = A_{(mn)}$ where $m\in \mathbb N$, I want to show that $(B_n)$ is a Markov chain with transition matrix $M^m$.

Please help me refine my argument. I have a picture of what is going on, but I don't know how to argue in symbols/more rigorously.

$B_n = A_{(mn)}$ means that you need to jump $m$ steps for each 1 jump in $A_n$ hence $(B_n)$ has the $m$-th transition matrix $M^m$. (How do I show that I can use $M$ as the "base transition matrix"?) Then to show that $(B_n) = Markov(\lambda, M^m)$, would I have to show that $\mathbb P(B_{n+1}=i_{n+1}|B_0=i_0,...,B_n=i_n)=(M^m)_{i_ni_{n+1}}$, or does that follow from the fact that $(A_n)$ is Markov?


share|cite|improve this question
Do you know Kolmogorov-Chapman equation? Will be of help. The structure you described is called a skeleton. See e.g. Meyn and Tweedie, 'Markov Chains and Stochastic Stability', p.67-68 – Ilya Oct 11 '11 at 11:31
@Gortaur: Thanks! – Henry D. Oct 12 '11 at 8:23
up vote 2 down vote accepted

Markov property for $(B_n)$ follows form the Markov property for $(A_n).$ To prove this write






$A_{mn}=i_n,A_{m(n-1)}=i_{n-1}\ldots,A_0=i_0)*\ldots*P(A_{mn+1}=j_1| A_{mn}=i_n,$


where $j_1,\ldots,j_{m-1}$ take all possible values from the state space. Then use Markov property of $(A_n)$: $P(B_{n+1}=i_{n+1}|B_n=i_n,\ldots,B_0=i_0)=\sum\limits_{j_1,\ldots,j_{m-1}}M_{i_nj_1}M_{j_1j_2}\ldots M_{j_{m-1}i_{n+1}}=M^m_{i_ni_{n+1}}.$

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.